Dysfunction of the basal ganglia has been associated with motor and cognitive deficits across a wide range of neurological and psychiatric disorders, including Parkinson's disease, obsessive compulsive disorder, and attention deficit disorder. The basal ganglia comprise two major circuits, the direct and indirect pathways, which are thought to have opposing effects on movement. While many previous studies examined how the balance of activity in these circuits shapes motor control, relatively little is known about their contributions to higher cognitive function. Recent work highlighted the basal ganglia's central role in reinforcement learning and decision-making and suggested that dysfunction in the decision process may underlie some of the cognitive deficits that are observed in basal ganglia disorders. One of the greatest barriers to progress is the fact that the direct and indirect pathwa projection neurons are intermingled and electrophysiologically indistinguishable, making it extremely difficult to study the functional role of these circuits in behaving animals. With the development of optogenetics, light-activated ion channels can be targeted to genetically defined neuronal populations, making it feasible to dissect the function of these circuits in behaving animals. Our goal is to define the pathway-specific mechanisms of decision making and reinforcement learning in the basal ganglia by applying advanced optical and electrophysiological methods. Recent studies using these tools revealed that the direct pathway plays a selective role in learning from reinforcement, and the indirect pathway plays a role in learning from punishment. This may have profound implications for drug addiction and depression, since selective deficits in learning from positive and negative outcomes may underlie some of the symptoms associated with these disorders. We will build on this work by investigating how pathway-specific neuronal activity and plasticity plays a role in learning in a dynamic decision-making task. With our findings we hope to further understand the basic circuit mechanisms underlying reinforcement learning, and to shed light on the higher cognitive deficits observed in basal ganglia disorders.